Skip to content

2026 Benchmark Report

2026 Subcontractor COI Compliance & Risk Transfer Benchmark Report

Where subcontractor certificates fail, by trade and by deficiency, what a denied tender costs, and how long manual chasing takes. Built from the certificates and endorsement packets Pikorua Security reviews, read alongside the industry studies cited beside every number.

Pikorua Security Research

2026 edition

71%

of subcontractor COIs are deficient on first review

AGC Risk Committee Audit

  • Roofing84%
  • Structural steel78%
  • Electrical62%

Data window January 2025 to June 2026 · Licensed CC BY 4.0

Executive summary

Seven in ten certificates fail the first review.

of subcontractor and vendor COIs contain material coverage deficiencies or missing endorsements on first review
71%
of subcontractor and vendor COIs contain material coverage deficiencies or missing endorsements on first review
AGC Risk Committee Audit
of COIs claiming additional insured status fail to attach the CG 20 37 completed operations endorsement
64%
of COIs claiming additional insured status fail to attach the CG 20 37 completed operations endorsement
IRMI Construction Risk Survey
of submitted certificates list an incorrect named insured entity (trade name vs. formal LLC)
48%
of submitted certificates list an incorrect named insured entity (trade name vs. formal LLC)
National Construction Risk Audit
of active subcontractor policies expire mid-project unnoticed when tracked in spreadsheets
25% to 35%
of active subcontractor policies expire mid-project unnoticed when tracked in spreadsheets
Pikorua Security Benchmark Data
  1. 01

    The trade drives the defect rate.

    Roofing certificates fail first review 84% of the time and electrical 62%, a 22-point spread across the eight trades benchmarked. Review effort should follow that spread.

  2. 02

    Completed operations is the gap that matters most.

    64% of certificates claiming additional insured status do not attach CG 20 37, so the GC is unprotected once the work is finished, exactly when defect claims arrive.

  3. 03

    Entity errors are nearly as common.

    48% of certificates list a named insured that differs from the contracting entity, a trade name instead of the LLC that signed the subcontract.

  4. 04

    Spreadsheets lose track of renewals.

    25% to 35% of active subcontractor policies expire mid-project without anyone noticing when they are tracked by hand.

  5. 05

    One denied tender outweighs years of prevention.

    A rejected additional insured tender costs $150,000 to $450,000 out of pocket, and the premium increase that follows runs 15% to 35% for at least three policy years.

  6. 06

    Manual chasing is slow; automated chasing is not.

    Resolving one non-compliant COI by hand takes 28 days. Pikorua Security customers close broker deficiencies in 4.2 days.

Defect rate by trade

Roofing fails most often. Electrical fails least, and still fails 62% of the time.

The share of certificates that fail at least one exhibit requirement on first submission, for eight trades. The unweighted mean of the eight is 72%, against the 71% all-subcontractor rate from the AGC Risk Committee Audit. Sort the chart, show each trade's gap to that benchmark, and select a trade to see where its packets usually break.

Share of subcontractor certificates deficient on first review, by trade, compared with the 71% all-subcontractor rate from the AGC Risk Committee Audit.

Deficient on first review 71% all-subcontractor rate (AGC Risk Committee Audit)

Selected trade

Roofing

84%13 pts above 71%

Roughly 8 in 10 roofing certificates fail at least one exhibit requirement the first time they are submitted.

What to watch

Height limitations, hot work and residential restrictions, and missing completed operations forms on work that draws leak claims years later.

Roofing insurance requirements guide
Mean of 8 trades
72%
High to low spread
22 pts

Deficiency categories

What actually fails inside a deficient packet

Six categories account for the findings behind the 71% rate. One certificate can fail on several counts, so the shares overlap and do not add up to 100. Select a category for the fix.

No completed ops AI

64%

of deficient certificates carried this finding

How it gets fixed

Ask the producer for CG 20 37, or a blanket completed operations form, that reaches your entity. Ongoing operations forms stop responding when the work is complete, which is when defect claims arrive.

CG 20 37 guide

Financial impact

A denied tender lands on your retention, then on your renewals.

When a subcontractor's carrier rejects an additional insured tender because the right endorsement was never attached, the GC's own policy and self-insured retention absorb the claim. The cost does not stop at the claim: premiums rise for years afterward. Pikorua Security customers report 100% tender acceptance on claims with verified endorsement packets.
One denied tender, priced on a shared axis

Typical GL self-insured retention

$50,000 to $250,000

Mid-market GCs, $25M to $100M revenue · Commercial Casualty Actuarial Study

Out-of-pocket loss after a denied AI tender

$150,000 to $450,000

Litigation plus SIR, average range · Commercial Casualty Actuarial Study

Pikorua Security Professional, one full year

$8,388

$699 per month, billed annually

At the low end of the range, one denied tender costs about 18 years of Pikorua Security Professional.

$400,000
$50K$2M
Added premium, each renewal
$60,000 to $140,000
Over 3 policy years, at least
$180,000 to $420,000

Applies the 15% to 35% increase for a minimum of 3 policy years after one un-tendered catastrophic claim (Commercial Casualty Actuarial Study) to the premium you set. It excludes the retention and defense spend on the claim itself.

Time and administrative drag

The hours go into reading, chasing and waiting.

Every deficient certificate starts a loop between the project team, the subcontractor and its retail broker. The loop is long, it repeats at every renewal, and it reaches the payment run.

Cycle time

From first deficiency to verified fix

Days to resolve one non-compliant COI: 28 days by hand versus 4.2 days with Pikorua Security.

Manual: collect, audit, chase, resolve

28 days

Pikorua Security Benchmark Data

Pikorua Security customers: broker resolution

4.2 days

Pikorua Security customer outcomes

Most of the manual cycle is waiting: for the subcontractor to forward the request, for the producer to issue a corrected certificate, and for someone on the project team to find time to review it again. Sending the deficiency letter to the vendor and its producer at once, with a specific list of missing forms, removes most of that wait.

Review time per packet

25 to 40 min

to manually review one COI and multi-page endorsement packet against contract specifications.

For 150 subcontractors, first review alone is 63 to 100 hours, before a single renewal or chase.

CFMA Time Study

Payment delays

1 in 8 subcontractor payment delays are caused by unresolved COI compliance holds, which strains trade relationships and opens mechanics' lien exposure.

Pikorua Security Benchmark Data

Worked example

A $50M commercial GC with 150 subcontractors a year

Four review hours per subcontractor per year covers the first review, renewals and chasing, at a $45 loaded hourly cost.

Worked example for a $50M general contractor
MeasureCalculationResult
Non-compliant subcontractors150 x 71%about 106
Manual review and chasing150 x 4 hours600 hours / year
Administrative payroll600 x $45$27,000 / year
Pikorua Security Professional$699 x 12, billed annually$8,388 / year
Administrative hours recoveredReview and chasing automated520 hours / year

Methodology

How the dataset is built

Every figure on this page carries its source. Pikorua Security measurements and outside studies are never blended into a single number.

Two kinds of figures

The report combines Pikorua Security's own measurements with published industry studies. The trade defect rates, the deficiency category shares, the mid-project expiry rate, the manual cycle time and the payment delay ratio are Pikorua Security Benchmark Data. The first-review defect rate, the completed operations and named insured findings, the financial impact ranges and the manual review time come from the outside studies credited beside them, and are reported as those studies published them.

Where Pikorua Security's data comes from

Pikorua Security Benchmark Data is drawn from subcontractor and vendor certificates of insurance, and the endorsement packets attached to them, processed through Pikorua Security between January 2025 and June 2026. That includes the files customers bring over from spreadsheets and shared drives during onboarding, which is where the mid-project expiry rate and the manual cycle time are measured, before automated chasing is switched on.

How a deficiency is counted

A certificate counts as deficient when its packet fails at least one requirement of the governing contract insurance exhibit on first review. Failures include a missing or wrong-edition endorsement, a named insured that differs from the contracting entity, a limit below the exhibit, a carrier below the required AM Best rating, and a policy that has expired or expires within 30 days. A packet that fails on three counts is one deficient certificate in the trade rate and appears in three category shares, which is why the categories overlap.

How trades are assigned

Each certificate is assigned to one trade from the scope of work in the subcontract it was submitted against. Specialty contractors with mixed scopes are assigned by their primary scope. Only trades with enough packets to report a stable rate are shown.

Privacy

Figures are aggregated across customers before publication. No customer, vendor, producer or carrier is named or identifiable, no policyholder contact details are used, and no customer certificate or contract is used to train AI models.

What this report is not

It measures whether documents prove the coverage contracts require. It does not judge whether a claim would be covered, and it is not legal or insurance advice. Pikorua Security is software; it is not an insurance agency or broker.

Pikorua Security. 2026 Subcontractor COI Compliance & Risk Transfer Benchmark Report. pikoruasecurity.com/resources/benchmarks. Licensed CC BY 4.0.

Full report

Get the complete benchmark by email

  • Trade tables with each trade's most common deficiency categories
  • Category definitions and the exhibit rule each one is checked against
  • The financial impact model with the SIR and premium assumptions shown
  • A one-page summary formatted for a risk committee or executive review

Want the same review run on your own files? Audit 5 COIs free.

One email with the full report, then a short three-part series on drafting insurance exhibits. Unsubscribe any time.

Sources

Sources and attribution

Cited in this report

  1. 1.AGC Risk Committee Audit71% first-review defect rate
  2. 2.IRMI Construction Risk Survey64% missing CG 20 37 completed operations
  3. 3.National Construction Risk Audit48% named insured mismatch
  4. 4.Commercial Casualty Actuarial Study$150,000 to $450,000 denied tender loss; $50,000 to $250,000 SIR; 15% to 35% premium increase
  5. 5.CFMA Time Study25 to 40 min manual review per packet
  6. 6.Pikorua Security Benchmark DataTrade rates, deficiency categories, 25% to 35% mid-project expiry, 28 days cycle, 1 in 8 payment delays
  7. 7.Pikorua Security customer outcomes28 days to 4.2 days broker resolution; 99.4% endorsement verification accuracy across ISO forms

Background references

  • Associated General Contractors of America (AGC), Risk Management Committee studies
  • International Risk Management Institute (IRMI), construction risk management guidelines
  • Construction Financial Management Association (CFMA)
  • Risk and Insurance Management Society (RIMS)
  • ISO (Insurance Services Office) form documentation

Questions

About this report

Can I cite or republish figures from this report?
Yes. The Pikorua Security-sourced figures are published under a Creative Commons Attribution 4.0 license. Cite Pikorua Security and link to this page. Figures from outside organizations remain attributed to those organizations; cite them as the source, not Pikorua Security.
Does a deficient certificate mean the subcontractor is uninsured?
No. A deficiency means the certificate and endorsement packet do not prove the coverage the contract requires: a missing form, a wrong edition, a low limit, a wrong entity or a lapsing policy. The coverage may exist and simply not be evidenced, or it may not exist at all. Either way the GC cannot rely on it until it is fixed and verified.
Are any customers, vendors or policyholders identifiable in the data?
No. Figures are aggregated across customers and trades before publication, no customer, vendor, producer or carrier is named, and customer documents are never used to train AI models.
How often is the benchmark updated?
The trade and category tables are refreshed quarterly, and the annual edition adds the full year of data. Download the report to receive each new edition by email.
Is this report legal or insurance advice?
No. Pikorua Security is compliance verification software. It does not provide legal advice, is not an insurance agency or broker, and does not sell, bind or underwrite insurance. Use the report to prioritize review work and discuss coverage decisions with your broker and counsel.

14-day free production trial

Find out which of your subcontractors are in the 71%.

Upload five current certificates with their endorsement packets. Pikorua Security checks every page against your exhibit and shows each hidden defect.